The AI Museum
AI Analytics1,072 recordsRecounted 23 September 2026

Closed-weight models in 2021

38

Models

21 organisations

0%

Open weights

0 open, 38 closed

7

Frontier

By Epoch's flag

1.2T

Largest traceable

GLaM

2021

Span

19 with a traceable size

Releases

by month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sept
Oct
Nov
Dec

Records

38
Closed-weight models in 2021, newest first. Each row links to the source Epoch AI records for it.
ModelOrganisationReleasedParametersWeights
ERNIE 3.0 TitanBaidu2021-12260BClosed
ERNIE-ViLGBaidu2021-1210B *Closed
GLaMGoogle2021-121.2TClosed
Gopher (280B)DeepMind2021-12280BClosed
Student of GamesDeepMind2021-12Closed
T-NLRv5 XXLMicrosoft2021-125BClosed
AliceMind-MMUAlibaba2021-11Closed
BASIC-LGoogle2021-113B *Closed
FlorenceMicrosoft2021-11893MClosed
NÜWAMicrosoft Research2021-11870MClosed
ViT-G/14 (LiT)Google Research2021-113BClosed
EfficientZeroTsinghua University2021-10Closed
EveHarvard Medical School2021-1015M *Closed
Megatron-Turing NLG 530BMicrosoft2021-10530BClosed
Yuan 1.0Inspur2021-10246BClosed
FLAN 137BGoogle Research2021-09137BClosed
HyperCLOVA 204BNAVER2021-09204B *Closed
MEBMicrosoft2021-09135BClosed
PermuteFormerPeking University2021-09150M *Closed
Turing ULRv5Microsoft2021-092BClosed
Jurassic-1-JumboAI21 Labs2021-08178BClosed
W2v-BERTGoogle Brain2021-081BClosed
CodexOpenAI2021-0712B *Closed
GOATDeepMind2021-073M *Closed
Adaptive Input Transformer + RDMicrosoft Research Asia2021-06247M *Closed
ALIGNGoogle Research2021-06820MClosed
CoAtNetGoogle2021-062BClosed
Fold2SeqIBM2021-0612M *Closed
ViT-G/14Google Brain2021-062BClosed
MedBERTPeng Cheng Laboratory2021-0517M *Closed
PLUGAlibaba2021-0427BClosed
M6-TAlibaba2021-031.0T *Closed
Meta Pseudo LabelsGoogle Brain2021-03480M *Closed
DLWPUniversity of Washington2021-023MClosed
Rational DQN AverageTU Darmstadt2021-022M *Closed
top-down frozen classifierUniversity of Edinburgh2021-02Closed
BigSSLGoogle2021-018B *Closed
DALL-EOpenAI2021-0112B *Closed

An asterisk marks a parameter figure the source does not consider traceable to the people who built the model. That judgement is the source’s and it has false negatives: Kimi K3 is marked speculative at 2.8T while Moonshot’s own model card states the figure. Why that distinction carries the whole exhibit.

Counts are of models Epoch AI records as notable, not of every model released. The most recent month is always incomplete, because a model enters the dataset when it is reviewed rather than when it ships.

Epoch AI, 'Data on AI Models'. Published online at epoch.ai. Retrieved from 'https://epoch.ai/data/ai-models-documentation' Used under CC BY 4.0.

How this museum is made